Date of Award


Degree Name

MS in Agriculture - Dairy Products Technology


Dairy Science


Rafael Jimenez-Flores


p.p1 {margin: 12.0px 0.0px 3.0px 0.0px; text-align: center; font: 16.0px 'Times New Roman'} p.p2 {margin: 0.0px 0.0px 10.0px 0.0px; text-align: center; font: 12.0px 'Times New Roman'} p.p3 {margin: 0.0px 0.0px 10.0px 0.0px; font: 12.0px 'Times New Roman'}


Combining Conventional Tests and Terminal Restriction Fragment Analysis to Evaluate Microbial Quality of Raw Milk

Haibin Guo

The dairy industry is an important part in the domestic economy in the U.S. and the quality of dairy products hinges on raw milk quality. Microorganisms play a critical role in raw milk quality and they are currently tested and monitored by conventional microbiological tests. Some of the most common conventional tests include somatic cell count (SCC), standard plate count (SPC), coliform count (CC), lab pasteurized count (LPC) and proteolytic strain count (PSC). However, these methods do not correlate with each other or with the quality of milk and milk products. One factor that contributes to this lack of correlation is the insufficient knowledge of microbial communities in raw milk. In this work we aimed to evaluate modern molecular methods to complement traditional quality procedures that may eventually complement conventional tests and improve milk quality evaluation. Therefore, a molecular method, Terminal Restriction Fragment (TRF) analysis was introduced. TRF analysis has been widely used as a tool to investigate the microbial communities in environmental samples. In this study, it was applied to investigate the microbial communities in raw milk and evaluate raw milk quality.

Milk samples were collected for over six months in the Cal Poly dairy farm and evaluated by conventional tests and TRF analysis. Samples were defined as “high quality” milk and “low quality” milk according to each conventional test first. The cutoffs applied were: 50,000 cfu/ml for SPC, 70,000 cells/ml for SCC, 100 cfu/ml for CC and 250 cfu/ml for LPC. TRF analysis was conducted on raw milk samples subsequently. DNA extraction was optimized. Non-Parametric Multivariate Analysis of Variance (NPMANOVA) was applied to TRF profiles from low and high quality milk. The analysis of Similarity of Percentage (SIMPER) was used to determine each TRF peak’s contribution to the dissimilarity between the profiles of high and low quality milk. The genus/species represented by TRF peaks were estimated via database matching. In addition, conventional tests and TRF analysis were also used to analyze the factors causing low quality milk. Rain event and cow’s apparent health were the two factors investigated since raw milk samples were collected from cows in different apparent health status on wet days and dry days.

Conventional tests revealed strong correlations between the results of SPC and PSC, and SPC and CC (coefficients of correlation > 0.8). It implied that the results of conventional tests might not be independent, so the statistics based on the assumption of independence of variables were not suitable to analyze the data. SCC showed no strong correlation with any other conventional tests. Raw milk samples were grouped as high quality and low quality according to SPC, CC, SCC and LPC. Using TRF analysis, it was found that there was a significant difference between TRF profiles from low and high quality milk when the quality was defined by SPC or LPC. A TRF peak at 268 bp generated by DpnII was predominant in the TRF profiles and had high abundance in the profiles of low quality milk. Hence, Pseudomonas spp. represented by TRF peak at 268 bp was likely the predominant bacteria in the microbial community associated with raw milk. TRF peaks at 61 bp, 81 bp, 104 bp, 104 bp, 201 bp, 242 bp, 268 bp, and 270 bp contributed the most to the dissimilarity between TRF profiles from different groups of samples. In addition, the presence of DNA derived from viable but non-culturable species that were associated with raw milk quality was detected.

Rain event was the most important factor affecting the microbial quality of raw milk in this study. Both the conventional tests and TRF analysis showed that there was a significant difference between raw milk samples collected on wet days versus dry days. Samples collected on wet days harbored high bacterial counts and had high abundance of the predominant TRF peaks. In addition, the same TRF peaks contributing the most to the dissimilarity between groups separated by rain event were found to be among those contributing the most to the dissimilarity between groups of high and low quality milk defined by conventional tests. During wet days, the low quality milk was likely caused by the increased dirtiness of cow’s teats. Soil microbes are often associated with microorganisms in raw milk such as psychrotrophic bacteria, coliform groups and spore-formers. Cow’s apparent health status showed no significant influence on the microbial quality of raw milk.

Overall, the combination of conventional tests and TRF analysis can yield a comprehensive understanding of microbial community in raw milk and improve the evaluation of raw milk quality. TRF analysis was demonstrated as a useful tool and a complement to conventional tests for milk quality evaluation by providing more information on the microbial community associated with raw milk. Findings in this study can offer a basis for further study and may help the dairy industry improve raw milk quality evaluation system.